From egoitz at sarenet.es Tue Dec 4 11:35:48 2018 From: egoitz at sarenet.es (Egoitz Aurrekoetxea) Date: Tue, 04 Dec 2018 17:35:48 +0100 Subject: Operating system performance for Cyrus and one more question Message-ID: Good afternoon, Is there any noticiable operating systems whose syscalls, filesystems... etc... could make Cyrus run faster?. We usually run it on FreeBSD and are pretty happy. Perhaps newer versions run better in any Linux like os?. About Xapian, XCONV, Jmap... I have tried doing a search with my mailbox in a new testing env, where I have connected a tesing mailbox imap partition with messages and mailboxes, coming from an older version of Cyrus. Even having Xapian, Conversations and so enabled, it does not find messages newer than 3 month located in the Sent folder. Perhaps Xapian, only works for new incoming messages?. Or should the mailboxes be reconstructed in order to a search to find and show all messages from a conversation existing in all mailboxes?. Or perhaps it's my mua's fault and is not searching properly in the IMAP server, because is not issuing some new commands to be used in order to make this kind of search to be done by Cyrus?. Best regards, -- EGOITZ AURREKOETXEA Departamento de sistemas 944 209 470 Parque Tecnol?gico. Edificio 103 48170 Zamudio (Bizkaia) egoitz at sarenet.es www.sarenet.es [1] Antes de imprimir este correo electr?nico piense si es necesario hacerlo. Links: ------ [1] http://www.sarenet.es -------------- next part -------------- An HTML attachment was scrubbed... URL: From jan.parcel at oracle.com Mon Dec 10 14:46:24 2018 From: jan.parcel at oracle.com (jan parcel) Date: Mon, 10 Dec 2018 11:46:24 -0800 Subject: Fwd: Re: [cyrusimap/cyrus-sasl] PAM auth causes segfault for unknown reason (#547) In-Reply-To: References: Message-ID: We have seen severe mutex issues with sasl_client init on super-fast machines. ?? We temporarily solved it by patching openldap to ensure it never calls sasl_client_init more than once,? but when I get out from under my current non-cyrus-sasl load, I had planned to test and submit changes to sasl_*_init to use a static mutext that is pre-initialized, which is what I found online as the solution for LIBRARY issues with passed in or even local mutexts. The issue is that the mutex needs a mutex, which needs a mutex, and it's turtles all the way down unless a static mutex is used at that top level. For some reason I have not understood, I've been told to not even THINK of doing anything like that to the mutexes used in gssapi. On 12/10/2018 7:53 AM, Alexander Sagen wrote: > > Trying to set up a temporary shared email server to replace an old > server with a dying disk while we move all our customers to a new > email solution. Configured saslauthd the same way it was configured on > the old server. > > Somehow, during the authentication process, saslauthd manages to > segfault, seemingly due to a mutex lock issue. > > Installed latest sasl2-bin (version 2.1.27) from APT (source: > http://eu-central-1.ec2.archive.ubuntu.com/ubuntu bionic/main amd64 > Packages). > > Running linux kernel 4.15.0-1021-aws. > > gdb output during crash: > > |root at mail1:~# gdb --args /usr/sbin/saslauthd -a pam -c -m > /var/spool/postfix/var/run/saslauthd -r -n 0 -d GNU gdb (Ubuntu > 8.1-0ubuntu3) 8.1.0.20180409-git Copyright (C) 2018 Free Software > Foundation, Inc. License GPLv3+: GNU GPL version 3 or later > This is free software: you are free > to change and redistribute it. There is NO WARRANTY, to the extent > permitted by law. Type "show copying" and "show warranty" for details. > This GDB was configured as "x86_64-linux-gnu". Type "show > configuration" for configuration details. For bug reporting > instructions, please see: . > Find the GDB manual and other documentation resources online at: > . For help, type > "help". Type "apropos word" to search for commands related to > "word"... Reading symbols from /usr/sbin/saslauthd...(no debugging > symbols found)...done. (gdb) run Starting program: /usr/sbin/saslauthd > -a pam -c -m /var/spool/postfix/var/run/saslauthd -r -n 0 -d [Thread > debugging using libthread_db enabled] Using host libthread_db library > "/lib/x86_64-linux-gnu/libthread_db.so.1". saslauthd[4767] :num_procs > : 0 saslauthd[4767] :mech_option: NULL saslauthd[4767] :run_path : > /var/spool/postfix/var/run/saslauthd saslauthd[4767] :auth_mech : pam > saslauthd[4767] :mmaped shared memory segment on file: > /var/spool/postfix/var/run/saslauthd/cache.mmap saslauthd[4767] > :bucket size: 96 bytes saslauthd[4767] :stats size : 36 bytes > saslauthd[4767] :timeout : 28800 seconds saslauthd[4767] :cache table: > 985828 total bytes saslauthd[4767] :cache table: 1711 slots > saslauthd[4767] :cache table: 10266 buckets saslauthd[4767] :flock > file opened at /var/spool/postfix/var/run/saslauthd/cache.flock > saslauthd[4767] :master pid is: 0 saslauthd[4767] :listening on > socket: /var/spool/postfix/var/run/saslauthd/mux saslauthd[4767] > :attempting a read lock on slot: 1501 saslauthd[4767] > :[login=someuser at example.com] [service=smtp] [realm=example.com]: not > found, update pending saslauthd[4767] :attempting to release lock on > slot: 1501 Program received signal SIGSEGV, Segmentation fault. > __GI___pthread_mutex_lock (mutex=0x20) at > ../nptl/pthread_mutex_lock.c:65 65 ../nptl/pthread_mutex_lock.c: No > such file or directory. (gdb) | > > saslauthd configuration: > > |START=yes DESC="SASL Authentication Daemon" NAME="saslauthd" > MECHANISMS="pam" MECH_OPTIONS="" THREADS=0 OPTIONS="-c -m > /var/spool/postfix/var/run/saslauthd -r" | > > ? > You are receiving this because you are subscribed to this thread. > Reply to this email directly, view it on GitHub > , or mute the > thread > . > -- Jan Parcel, Software Developer Oracle Systems Server & Cloud Engineering -------------- next part -------------- An HTML attachment was scrubbed... URL: From dilyan.palauzov at aegee.org Tue Dec 11 08:45:24 2018 From: dilyan.palauzov at aegee.org (=?UTF-8?Q?=D0=94=D0=B8=D0=BB=D1=8F=D0=BD_?= =?UTF-8?Q?=D0=9F=D0=B0=D0=BB=D0=B0=D1=83=D0=B7=D0=BE=D0=B2?=) Date: Tue, 11 Dec 2018 13:45:24 +0000 Subject: Fwd: Re: [cyrusimap/cyrus-sasl] PAM auth causes segfault for unknown reason (#547) In-Reply-To: References: Message-ID: Hello, the message below does not show the full backtrace at the very last moment. One way to create it is to produce a core dump and then open with gdb the saslauthd executable, the coredump and call "thread apply all bt f". More usefull could be to run the process under valgrind/memchech, after compiling it debug information. Otherwise it is evident, that the process terminates unexpectedly, but it is unclear why and where. Regards ????? On Mon, 2018-12-10 at 11:46 -0800, jan parcel wrote: > We have seen severe mutex issues with sasl_client init on super-fast machines. We temporarily solved it by patching openldap to ensure it never calls sasl_client_init more than once, but when I get out from under my current non-cyrus-sasl load, I had planned to test and submit changes to sasl_*_init to use a static mutext that is pre-initialized, which is what I found online as the solution for LIBRARY issues with passed in or even local mutexts. > > The issue is that the mutex needs a mutex, which needs a mutex, and it's turtles all the way down unless a static mutex is used at that top level. > > For some reason I have not understood, I've been told to not even THINK of doing anything like that to the mutexes used in gssapi. > > > On 12/10/2018 7:53 AM, Alexander Sagen wrote: > > Trying to set up a temporary shared email server to replace an old server with a dying disk while we move all our customers to a new email solution. Configured saslauthd the same way it was configured on the old server. > > Somehow, during the authentication process, saslauthd manages to segfault, seemingly due to a mutex lock issue. > > Installed latest sasl2-bin (version 2.1.27) from APT (source: http://eu-central-1.ec2.archive.ubuntu.com/ubuntu bionic/main amd64 Packages). > > Running linux kernel 4.15.0-1021-aws. > > gdb output during crash: > > root at mail1:~# gdb --args /usr/sbin/saslauthd -a pam -c -m /var/spool/postfix/var/run/saslauthd -r -n 0 -d > > GNU gdb (Ubuntu 8.1-0ubuntu3) 8.1.0.20180409-git > > Copyright (C) 2018 Free Software Foundation, Inc. > > License GPLv3+: GNU GPL version 3 or later > > This is free software: you are free to change and redistribute it. > > There is NO WARRANTY, to the extent permitted by law. Type "show copying" > > and "show warranty" for details. > > This GDB was configured as "x86_64-linux-gnu". > > Type "show configuration" for configuration details. > > For bug reporting instructions, please see: > > ;. > > Find the GDB manual and other documentation resources online at: > > ;. > > For help, type "help". > > Type "apropos word" to search for commands related to "word"... > > Reading symbols from /usr/sbin/saslauthd...(no debugging symbols found)...done. > > (gdb) run > > Starting program: /usr/sbin/saslauthd -a pam -c -m /var/spool/postfix/var/run/saslauthd -r -n 0 -d > > [Thread debugging using libthread_db enabled] > > Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1". > > saslauthd[4767] :num_procs : 0 > > saslauthd[4767] :mech_option: NULL > > saslauthd[4767] :run_path : /var/spool/postfix/var/run/saslauthd > > saslauthd[4767] :auth_mech : pam > > saslauthd[4767] :mmaped shared memory segment on file: /var/spool/postfix/var/run/saslauthd/cache.mmap > > saslauthd[4767] :bucket size: 96 bytes > > saslauthd[4767] :stats size : 36 bytes > > saslauthd[4767] :timeout : 28800 seconds > > saslauthd[4767] :cache table: 985828 total bytes > > saslauthd[4767] :cache table: 1711 slots > > saslauthd[4767] :cache table: 10266 buckets > > saslauthd[4767] :flock file opened at /var/spool/postfix/var/run/saslauthd/cache.flock > > saslauthd[4767] :master pid is: 0 > > saslauthd[4767] :listening on socket: /var/spool/postfix/var/run/saslauthd/mux > > saslauthd[4767] :attempting a read lock on slot: 1501 > > saslauthd[4767] :[login=someuser at example.com] [service=smtp] [realm=example.com]: not found, update pending > > saslauthd[4767] :attempting to release lock on slot: 1501 > > > > Program received signal SIGSEGV, Segmentation fault. > > __GI___pthread_mutex_lock (mutex=0x20) at ../nptl/pthread_mutex_lock.c:65 > > 65 ../nptl/pthread_mutex_lock.c: No such file or directory. > > (gdb) > > saslauthd configuration: > > START=yes > > DESC="SASL Authentication Daemon" > > NAME="saslauthd" > > MECHANISMS="pam" > > MECH_OPTIONS="" > > THREADS=0 > > OPTIONS="-c -m /var/spool/postfix/var/run/saslauthd -r" > > ? > > You are receiving this because you are subscribed to this thread. > > Reply to this email directly, view it on GitHub, or mute the thread. > From egoitz at sarenet.es Thu Dec 13 10:52:28 2018 From: egoitz at sarenet.es (Egoitz Aurrekoetxea) Date: Thu, 13 Dec 2018 16:52:28 +0100 Subject: Question for upgrading Message-ID: <9e71e82ddcba5500a175c0e447131413@sarenet.es> Good afternoon, I was trying to upgrade part of our Cyrus imap installation, concretely that one consisting in still 2.3. I was planning to set up Cyrus 3.0. I have seen all works properly except for the unexpunge command because as someone stated here, a reconstruct -V max was needed.The problem is that this reconstruct command, takes ages and I'm not able to keep the service offline so many time. So I have been thinking in the following scenario : - Cyrus 2.3 master -> Cyrus 2.4 slave Get this 2.4 slave ready and set it as master. But here comes my first doubt. Does the 2.4 replication work with the 2.3 replication?. Can in this pair, both (the 2.3 and the 2.4) be both master and slave?. I mean to switch roles in the pair. Make one become master and the other slave and vice versa?. Let's think now Cyrus 2.4 is ready and working. - Now, I would set up a new 3.0 slave. I know 2.4 could replicate with 3.0. So I would get the 3.0 ready and then set 3.0 as master. Can in this pair both the 2.4 and 3.0 be master and slave?. Meaning again to the same role switching commented before... to make one to be master and the other slave or vice versa.... I'l will end up with 2 3.0 master and slave... but I need to trace the path... Does anyone see any other way?. Best regards, - -- EGOITZ AURREKOETXEA Departamento de sistemas 944 209 470 Parque Tecnol?gico. Edificio 103 48170 Zamudio (Bizkaia) egoitz at sarenet.es www.sarenet.es [1] Antes de imprimir este correo electr?nico piense si es necesario hacerlo. Links: ------ [1] http://www.sarenet.es -------------- next part -------------- An HTML attachment was scrubbed... URL: From egoitz at sarenet.es Thu Dec 13 13:25:40 2018 From: egoitz at sarenet.es (Egoitz Aurrekoetxea) Date: Thu, 13 Dec 2018 19:25:40 +0100 Subject: Question for upgrading In-Reply-To: <9e71e82ddcba5500a175c0e447131413@sarenet.es> References: <9e71e82ddcba5500a175c0e447131413@sarenet.es> Message-ID: <5ed0a3707ca4b486ede31779d1cbf4da@sarenet.es> Hi again! Else as a simplication way can you replicate any manner a 2.3 with some newer version?. At least in manual mode (not rolling)?. Cheers. --- EGOITZ AURREKOETXEA Departamento de sistemas 944 209 470 Parque Tecnol?gico. Edificio 103 48170 Zamudio (Bizkaia) egoitz at sarenet.es www.sarenet.es [1] Antes de imprimir este correo electr?nico piense si es necesario hacerlo. El 13-12-2018 16:52, Egoitz Aurrekoetxea escribi?: > Good afternoon, > > I was trying to upgrade part of our Cyrus imap installation, concretely that one consisting in still 2.3. I was planning to set up Cyrus 3.0. I have seen all works properly except for the unexpunge command because as someone stated here, a reconstruct -V max was needed.The problem is that this reconstruct command, takes ages and I'm not able to keep the service offline so many time. So I have been thinking in the following scenario : > > - Cyrus 2.3 master -> Cyrus 2.4 slave > > Get this 2.4 slave ready and set it as master. But here comes my first doubt. Does the 2.4 replication work with the 2.3 replication?. Can in this pair, both (the 2.3 and the 2.4) be both master and slave?. I mean to switch roles in the pair. Make one become master and the other slave and vice versa?. > > Let's think now Cyrus 2.4 is ready and working. > > - Now, I would set up a new 3.0 slave. I know 2.4 could replicate with 3.0. So I would get the 3.0 ready and then set 3.0 as master. Can in this pair both the 2.4 and 3.0 be master and slave?. Meaning again to the same role switching commented before... to make one to be master and the other slave or vice versa.... > > I'l will end up with 2 3.0 master and slave... but I need to trace the path... > > Does anyone see any other way?. > > Best regards, > > - > > -- > > EGOITZ AURREKOETXEA > Departamento de sistemas > 944 209 470 > Parque Tecnol?gico. Edificio 103 > 48170 Zamudio (Bizkaia) > egoitz at sarenet.es > www.sarenet.es [1] > Antes de imprimir este correo electr?nico piense si es necesario hacerlo. > ---- > Cyrus Home Page: http://www.cyrusimap.org/ > List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ > To Unsubscribe: > https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus Links: ------ [1] http://www.sarenet.es -------------- next part -------------- An HTML attachment was scrubbed... URL: From egoitz at sarenet.es Fri Dec 14 10:11:51 2018 From: egoitz at sarenet.es (Egoitz Aurrekoetxea) Date: Fri, 14 Dec 2018 16:11:51 +0100 Subject: Question for upgrading In-Reply-To: <5ed0a3707ca4b486ede31779d1cbf4da@sarenet.es> References: <9e71e82ddcba5500a175c0e447131413@sarenet.es> <5ed0a3707ca4b486ede31779d1cbf4da@sarenet.es> Message-ID: <2715a40ab7c773b132308983d1aed5a0@sarenet.es> Common Bron, Ellie, mates :) any advise? Cheers :) :) --- EGOITZ AURREKOETXEA Departamento de sistemas 944 209 470 Parque Tecnol?gico. Edificio 103 48170 Zamudio (Bizkaia) egoitz at sarenet.es www.sarenet.es [1] Antes de imprimir este correo electr?nico piense si es necesario hacerlo. El 13-12-2018 19:25, Egoitz Aurrekoetxea escribi?: > Hi again! > > Else as a simplication way can you replicate any manner a 2.3 with some newer version?. At least in manual mode (not rolling)?. > > Cheers. > > --- > > EGOITZ AURREKOETXEA > Departamento de sistemas > 944 209 470 > Parque Tecnol?gico. Edificio 103 > 48170 Zamudio (Bizkaia) > egoitz at sarenet.es > www.sarenet.es [1] > Antes de imprimir este correo electr?nico piense si es necesario hacerlo. > > El 13-12-2018 16:52, Egoitz Aurrekoetxea escribi?: > >> Good afternoon, >> >> I was trying to upgrade part of our Cyrus imap installation, concretely that one consisting in still 2.3. I was planning to set up Cyrus 3.0. I have seen all works properly except for the unexpunge command because as someone stated here, a reconstruct -V max was needed.The problem is that this reconstruct command, takes ages and I'm not able to keep the service offline so many time. So I have been thinking in the following scenario : >> >> - Cyrus 2.3 master -> Cyrus 2.4 slave >> >> Get this 2.4 slave ready and set it as master. But here comes my first doubt. Does the 2.4 replication work with the 2.3 replication?. Can in this pair, both (the 2.3 and the 2.4) be both master and slave?. I mean to switch roles in the pair. Make one become master and the other slave and vice versa?. >> >> Let's think now Cyrus 2.4 is ready and working. >> >> - Now, I would set up a new 3.0 slave. I know 2.4 could replicate with 3.0. So I would get the 3.0 ready and then set 3.0 as master. Can in this pair both the 2.4 and 3.0 be master and slave?. Meaning again to the same role switching commented before... to make one to be master and the other slave or vice versa.... >> >> I'l will end up with 2 3.0 master and slave... but I need to trace the path... >> >> Does anyone see any other way?. >> >> Best regards, >> >> - >> >> -- >> >> EGOITZ AURREKOETXEA >> Departamento de sistemas >> 944 209 470 >> Parque Tecnol?gico. Edificio 103 >> 48170 Zamudio (Bizkaia) >> egoitz at sarenet.es >> www.sarenet.es [1] >> Antes de imprimir este correo electr?nico piense si es necesario hacerlo. >> ---- >> Cyrus Home Page: http://www.cyrusimap.org/ >> List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ >> To Unsubscribe: >> https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus > > ---- > Cyrus Home Page: http://www.cyrusimap.org/ > List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ > To Unsubscribe: > https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus Links: ------ [1] http://www.sarenet.es -------------- next part -------------- An HTML attachment was scrubbed... URL: From nic at onlight.com Sat Dec 15 11:05:29 2018 From: nic at onlight.com (Nic Bernstein) Date: Sat, 15 Dec 2018 10:05:29 -0600 Subject: Question for upgrading In-Reply-To: <9e71e82ddcba5500a175c0e447131413@sarenet.es> References: <9e71e82ddcba5500a175c0e447131413@sarenet.es> Message-ID: <07899a02-1bc4-8f47-8f47-7cc0d4f3f7ee@onlight.com> On 12/13/18 9:52 AM, Egoitz Aurrekoetxea wrote: > > I was trying to upgrade part of our Cyrus imap installation, > concretely that one consisting in still 2.3. I was planning to set up > Cyrus 3.0. I have seen all works properly except for the unexpunge > command because as someone stated here, a reconstruct -V max was > needed.The problem is that this reconstruct command, takes ages and > I'm not able to keep the service offline so many time. So I have been > thinking in the following scenario : > Egoitz, As long as you've followed all of the various steps needed to account for version changes, as outlined in the Release Notes for /all/ intermediary releases, then the last step should be the updating of the indexes.? Rather than running "reconstruct -V max" on all mailboxes at once, simply run it on subsets.? We've done this on large installations without ill effect.? You can have the service on line during the reconstruct, and, as you note, have most all functionality present. We build a list of users (mboxlist.txt), and then run a cron job, during off-peak hours, using the 'parallel' command like so (from 'crontab -e -u cyrus'): ### ## Start a batch of recursive reconstruct jobs at 7PM and discontinue ## at 6:53AM. 0 19 * * * parallel -j 5 --resume --joblog /var/lib/imap/reconstruct.log cyrus reconstruct -G -V max -r -u {} < /var/lib/imap/mboxlist.txt 53 06 * * * kill -TERM `ps ax | grep [p]arallel | awk '{print $1}'` Cheers, ??? -nic -- Nic Bernstein nic at onlight.com Onlight Inc. www.onlight.com 6525 W Bluemound Rd., Ste 24 v. 414.272.4477 Milwaukee, Wisconsin 53213-4073 f. 414.290.0335 -------------- next part -------------- An HTML attachment was scrubbed... URL: From jc at irbs.com Sun Dec 16 16:24:35 2018 From: jc at irbs.com (John Capo) Date: Sun, 16 Dec 2018 16:24:35 -0500 (EST) Subject: Question for upgrading In-Reply-To: <5ed0a3707ca4b486ede31779d1cbf4da@sarenet.es> References: <9e71e82ddcba5500a175c0e447131413@sarenet.es> <5ed0a3707ca4b486ede31779d1cbf4da@sarenet.es> Message-ID: <2063.205.237.194.2.1544995475.squirrel@beta.mxes.net> On Thu, December 13, 2018 13:25, Egoitz Aurrekoetxea wrote: > Hi again! > > > Else as a simplication way can you replicate any manner a 2.3 with some > newer version?. At least in manual mode (not rolling)?. The replication protocol in 2.4 is not compatible wit 2.3. There is no easy way. I'm rsync'ing one account at a time between 2.3 and 2.4 with IMAP/POP and LMTP access disabled, followed by a reconstruct, and then enabling IMAP/POP and LMTP access. Many terrabytes to go. You could also look at imapsync but its slow. > > Cheers. > > > --- > > > EGOITZ AURREKOETXEA > Departamento de sistemas > 944 209 470 > Parque Tecnol??gico. Edificio 103 > 48170 Zamudio (Bizkaia) > egoitz at sarenet.es www.sarenet.es [1] Antes de imprimir este correo electr??nico piense si es > necesario hacerlo. > > El 13-12-2018 16:52, Egoitz Aurrekoetxea escribi??: > > >> Good afternoon, >> >> >> I was trying to upgrade part of our Cyrus imap installation, concretely that one consisting in >> still 2.3. I was planning to set up Cyrus 3.0. I have seen all works properly except for the >> unexpunge command because as someone stated here, a reconstruct -V max was needed.The problem >> is that this reconstruct command, takes ages and I'm not able to keep the service offline so >> many time. So I have been thinking in the following scenario : >> >> - Cyrus 2.3 master -> Cyrus 2.4 slave >> >> >> Get this 2.4 slave ready and set it as master. But here comes my first doubt. Does the 2.4 >> replication work with the 2.3 replication?. Can in this pair, both (the 2.3 and the 2.4) be >> both master and slave?. I mean to switch roles in the pair. Make one become master and the >> other slave and vice versa?. >> >> Let's think now Cyrus 2.4 is ready and working. >> >> >> - Now, I would set up a new 3.0 slave. I know 2.4 could replicate with 3.0. So I would get the >> 3.0 ready and then set 3.0 as master. Can in this pair both the 2.4 and 3.0 be master and >> slave?. Meaning again to the same role switching commented before... to make one to be master >> and the other slave or vice versa.... >> >> I'l will end up with 2 3.0 master and slave... but I need to trace the path... >> >> >> Does anyone see any other way?. >> >> >> Best regards, >> >> >> - >> >> >> -- >> >> >> EGOITZ AURREKOETXEA >> Departamento de sistemas >> 944 209 470 >> Parque Tecnol??gico. Edificio 103 >> 48170 Zamudio (Bizkaia) >> egoitz at sarenet.es www.sarenet.es [1] Antes de imprimir este correo electr??nico piense si es >> necesario hacerlo. ---- >> Cyrus Home Page: http://www.cyrusimap.org/ >> List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ >> To Unsubscribe: >> https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus >> > > > Links: > ------ > [1] http://www.sarenet.es From egoitz at sarenet.es Mon Dec 17 04:40:36 2018 From: egoitz at sarenet.es (Egoitz Aurrekoetxea) Date: Mon, 17 Dec 2018 10:40:36 +0100 Subject: Question for upgrading In-Reply-To: <2063.205.237.194.2.1544995475.squirrel@beta.mxes.net> References: <9e71e82ddcba5500a175c0e447131413@sarenet.es> <5ed0a3707ca4b486ede31779d1cbf4da@sarenet.es> <2063.205.237.194.2.1544995475.squirrel@beta.mxes.net> Message-ID: <38a2eed166ef29ab1f38cd21660465a0@sarenet.es> Hi mate, I think (and say think :) )I finally found a method. Although I'm testing it deeply... it seems (say seems too :) ) 2.4 is compatible with a mail spool in 2.3 (at least with my config). So I'll try to upgrade first to 2.4 and later to 3.0 setting up a replication from 2.4 to 3.0. Would be fine if Bron, Ellie or someone at Fastmail could tell something about it to us :) :) Cheers! --- EGOITZ AURREKOETXEA Departamento de sistemas 944 209 470 Parque Tecnol?gico. Edificio 103 48170 Zamudio (Bizkaia) egoitz at sarenet.es www.sarenet.es [1] Antes de imprimir este correo electr?nico piense si es necesario hacerlo. El 16-12-2018 22:24, John Capo escribi?: > On Thu, December 13, 2018 13:25, Egoitz Aurrekoetxea wrote: > >> Hi again! >> >> Else as a simplication way can you replicate any manner a 2.3 with some >> newer version?. At least in manual mode (not rolling)?. > > The replication protocol in 2.4 is not compatible wit 2.3. > > There is no easy way. I'm rsync'ing one account at a time between 2.3 and 2.4 with IMAP/POP and LMTP access disabled, followed by a reconstruct, and then enabling IMAP/POP and LMTP access. > > Many terrabytes to go. > > You could also look at imapsync but its slow. > > Cheers. > > --- > > EGOITZ AURREKOETXEA > Departamento de sistemas > 944 209 470 > Parque Tecnol??gico. Edificio 103 > 48170 Zamudio (Bizkaia) > egoitz at sarenet.es www.sarenet.es [1] [1 [1]] Antes de imprimir este correo electr??nico piense si es > necesario hacerlo. > > El 13-12-2018 16:52, Egoitz Aurrekoetxea escribi??: > > Good afternoon, > > I was trying to upgrade part of our Cyrus imap installation, concretely that one consisting in > still 2.3. I was planning to set up Cyrus 3.0. I have seen all works properly except for the > unexpunge command because as someone stated here, a reconstruct -V max was needed.The problem > is that this reconstruct command, takes ages and I'm not able to keep the service offline so > many time. So I have been thinking in the following scenario : > > - Cyrus 2.3 master -> Cyrus 2.4 slave > > Get this 2.4 slave ready and set it as master. But here comes my first doubt. Does the 2.4 > replication work with the 2.3 replication?. Can in this pair, both (the 2.3 and the 2.4) be > both master and slave?. I mean to switch roles in the pair. Make one become master and the > other slave and vice versa?. > > Let's think now Cyrus 2.4 is ready and working. > > - Now, I would set up a new 3.0 slave. I know 2.4 could replicate with 3.0. So I would get the > 3.0 ready and then set 3.0 as master. Can in this pair both the 2.4 and 3.0 be master and > slave?. Meaning again to the same role switching commented before... to make one to be master > and the other slave or vice versa.... > > I'l will end up with 2 3.0 master and slave... but I need to trace the path... > > Does anyone see any other way?. > > Best regards, > > - > > -- > > EGOITZ AURREKOETXEA > Departamento de sistemas > 944 209 470 > Parque Tecnol??gico. Edificio 103 > 48170 Zamudio (Bizkaia) > egoitz at sarenet.es www.sarenet.es [1] [1 [1]] Antes de imprimir este correo electr??nico piense si es > necesario hacerlo. ---- > Cyrus Home Page: http://www.cyrusimap.org/ > List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ > To Unsubscribe: > https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus > > Links: > ------ > [1] http://www.sarenet.es Links: ------ [1] http://www.sarenet.es -------------- next part -------------- An HTML attachment was scrubbed... URL: From rsto at fastmailteam.com Mon Dec 17 16:27:59 2018 From: rsto at fastmailteam.com (Robert Stepanek) Date: Mon, 17 Dec 2018 22:27:59 +0100 Subject: Notes Dec 17, 2018 Message-ID: <1545082079.1957914.1611824040.3B4247C4@webmail.messagingengine.com> Participants: Ken, Ellie, Bron, Rik, Partha, RobS, Yves Ken: Working on mailbox unique ids. All user-metadata now keyed by unique id of user inbox. Current work is on annotations.db, some Cassandane tests failing. Next are conversations, hopefully changing F records should be enough. Bron: has patch for intermediate mailboxes to clean parents where all children are intermediates. Rather expensive during startup O(n) for depth of tree, haven't tried on a large server. Fixed also a transaction db error. JMAP: count of non-deleted emails is bogus currently due to delayed expunge. Might need to store count. Will look into it, and if it requires more work will hand over to RobS. RobS: QUOTAMODSEQ and RACLMODSEQ on master, and in production. Haven't got issue reports, looks clean. Currently working on patch for Cyrus DB read-only locks for conversations.db. Will run full tests tomorrow, then ask for code review and to deploy in test environment with more concurrency. With Google Hangouts getting sunsetted, we are looking for a new video conferencing tool for our meetings. Will try appear.in. No Cyrus meeting scheduled over Christmas and New Years Eve, so next regular Cyrus meeting will be on January 7, 2019. -------------- next part -------------- An HTML attachment was scrubbed... URL: From ellie at fastmail.com Mon Dec 17 16:46:06 2018 From: ellie at fastmail.com (ellie timoney) Date: Tue, 18 Dec 2018 08:46:06 +1100 Subject: Question for upgrading In-Reply-To: <38a2eed166ef29ab1f38cd21660465a0@sarenet.es> References: <9e71e82ddcba5500a175c0e447131413@sarenet.es> <5ed0a3707ca4b486ede31779d1cbf4da@sarenet.es> <2063.205.237.194.2.1544995475.squirrel@beta.mxes.net> <38a2eed166ef29ab1f38cd21660465a0@sarenet.es> Message-ID: <1545083166.1800663.1611834480.2001673C@webmail.messagingengine.com> > Would be fine if Bron, Ellie or someone at Fastmail could tell > something about it to us :) :) Cheeky! Thing is, at FM our production servers more or less track the master branch, so we were running "3.0" from the moment 2.5 forked off, and so our upgrade process only ever really needs to deal with iterative changes. Our accumulated experience is what informs the "upgrade notes" included with the 2.5.0 and 3.0.0 releases (not sure about older versions, I wasn't around then). This mailing list is the right place to find people with real experience doing big-bang upgrades. :) Cheers, ellie On Mon, Dec 17, 2018, at 8:40 PM, Egoitz Aurrekoetxea wrote: > Hi mate, > > I think (and say think :) )I finally found a method. Although I'm > testing it deeply... it seems (say seems too :) ) 2.4 is compatible > with a mail spool in 2.3 (at least with my config). So I'll try to > upgrade first to 2.4 and later to 3.0 setting up a replication from > 2.4 to 3.0. Would be fine if Bron, Ellie or someone at Fastmail could > tell something about it to us :) :)> > Cheers! > --- > > sarenet > *Egoitz Aurrekoetxea* > Departamento de sistemas > 944 209 470 > Parque Tecnol?gico. Edificio 103 > 48170 Zamudio (Bizkaia) > egoitz at sarenet.es > www.sarenet.es > > Antes de imprimir este correo electr?nico piense si es necesario > hacerlo.> > El 16-12-2018 22:24, John Capo escribi?: >> On Thu, December 13, 2018 13:25, Egoitz Aurrekoetxea wrote: >>> Hi again! >>> >>> >>> Else as a simplication way can you replicate any manner a 2.3 >>> with some>>> newer version?. At least in manual mode (not rolling)?. >> >> The replication protocol in 2.4 is not compatible wit 2.3. >> >> There is no easy way. I'm rsync'ing one account at a time between >> 2.3 and 2.4 with IMAP/POP and LMTP access disabled, followed by a >> reconstruct, and then enabling IMAP/POP and LMTP access.>> >> Many terrabytes to go. >> >> You could also look at imapsync but its slow. >> >> >>> >>> Cheers. >>> >>> >>> --- >>> >>> >>> EGOITZ AURREKOETXEA >>> Departamento de sistemas >>> 944 209 470 >>> Parque Tecnol??gico. Edificio 103 >>> 48170 Zamudio (Bizkaia) >>> egoitz at sarenet.es www.sarenet.es [1[1]] Antes de imprimir este >>> correo electr??nico piense si es>>> necesario hacerlo. >>> >>> El 13-12-2018 16:52, Egoitz Aurrekoetxea escribi??: >>> >>> >>> >>>> Good afternoon, >>>> >>>> >>>> I was trying to upgrade part of our Cyrus imap installation, >>>> concretely that one consisting in>>>> still 2.3. I was planning to set up Cyrus 3.0. I have seen all >>>> works properly except for the>>>> unexpunge command because as someone stated here, a reconstruct -V >>>> max was needed.The problem>>>> is that this reconstruct command, takes ages and I'm not able to >>>> keep the service offline so>>>> many time. So I have been thinking in the following scenario : >>>> >>>> - Cyrus 2.3 master -> Cyrus 2.4 slave >>>> >>>> >>>> Get this 2.4 slave ready and set it as master. But here comes my >>>> first doubt. Does the 2.4>>>> replication work with the 2.3 replication?. Can in this pair, both >>>> (the 2.3 and the 2.4) be>>>> both master and slave?. I mean to switch roles in the pair. Make >>>> one become master and the>>>> other slave and vice versa?. >>>> >>>> Let's think now Cyrus 2.4 is ready and working. >>>> >>>> >>>> - Now, I would set up a new 3.0 slave. I know 2.4 could replicate >>>> with 3.0. So I would get the>>>> 3. ready and then set 3.0 as master. Can in this pair both the 2.4 >>>> and 3.0 be master and>>>> slave?. Meaning again to the same role switching commented >>>> before... to make one to be master>>>> and the other slave or vice versa.... >>>> >>>> I'l will end up with 2 3.0 master and slave... but I need to trace >>>> the path...>>>> >>>> >>>> Does anyone see any other way?. >>>> >>>> >>>> Best regards, >>>> >>>> >>>> - >>>> >>>> >>>> -- >>>> >>>> >>>> EGOITZ AURREKOETXEA >>>> Departamento de sistemas >>>> 944 209 470 >>>> Parque Tecnol??gico. Edificio 103 >>>> 48170 Zamudio (Bizkaia) >>>> egoitz at sarenet.es www.sarenet.es [1[2]] Antes de imprimir este >>>> correo electr??nico piense si es>>>> necesario hacerlo. ---- >>>> Cyrus Home Page: http://www.cyrusimap.org/ >>>> List Archives/Info: >>>> http://lists.andrew.cmu.edu/pipermail/info-cyrus/>>>> To Unsubscribe: >>>> https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus >>> >>> >>> Links: >>> ------ >>> [1] http://www.sarenet.es Links: 1. http://www.sarenet.es 2. http://www.sarenet.es -------------- next part -------------- An HTML attachment was scrubbed... URL: From egoitz at sarenet.es Tue Dec 18 01:26:24 2018 From: egoitz at sarenet.es (Egoitz Aurrekoetxea) Date: Tue, 18 Dec 2018 07:26:24 +0100 Subject: Question for upgrading In-Reply-To: <1545083166.1800663.1611834480.2001673C@webmail.messagingengine.com> References: <9e71e82ddcba5500a175c0e447131413@sarenet.es> <5ed0a3707ca4b486ede31779d1cbf4da@sarenet.es> <2063.205.237.194.2.1544995475.squirrel@beta.mxes.net> <38a2eed166ef29ab1f38cd21660465a0@sarenet.es> <1545083166.1800663.1611834480.2001673C@webmail.messagingengine.com> Message-ID: <0f434203c44493a197ec5aa94d5267ac@sarenet.es> Thank you so much Ellie!! I would test, read code when needed and finally attempt the upgrade... I think this coud work.... should say I'm in testing phase still, I'll keep you informed :) Cheers! --- EGOITZ AURREKOETXEA Departamento de sistemas 944 209 470 Parque Tecnol?gico. Edificio 103 48170 Zamudio (Bizkaia) egoitz at sarenet.es www.sarenet.es [1] Antes de imprimir este correo electr?nico piense si es necesario hacerlo. El 17-12-2018 22:46, ellie timoney escribi?: >> Would be fine if Bron, Ellie or someone at Fastmail could tell something about it to us :) :) > > Cheeky! Thing is, at FM our production servers more or less track the master branch, so we were running "3.0" from the moment 2.5 forked off, and so our upgrade process only ever really needs to deal with iterative changes. Our accumulated experience is what informs the "upgrade notes" included with the 2.5.0 and 3.0.0 releases (not sure about older versions, I wasn't around then). > > This mailing list is the right place to find people with real experience doing big-bang upgrades. :) > > Cheers, > > ellie > > On Mon, Dec 17, 2018, at 8:40 PM, Egoitz Aurrekoetxea wrote: > > Hi mate, > > I think (and say think :) )I finally found a method. Although I'm testing it deeply... it seems (say seems too :) ) 2.4 is compatible with a mail spool in 2.3 (at least with my config). So I'll try to upgrade first to 2.4 and later to 3.0 setting up a replication from 2.4 to 3.0. Would be fine if Bron, Ellie or someone at Fastmail could tell something about it to us :) :) > > Cheers! > > --- > > EGOITZ AURREKOETXEA > Departamento de sistemas > > 944 209 470 > Parque Tecnol?gico. Edificio 103 > 48170 Zamudio (Bizkaia) > egoitz at sarenet.es > > www.sarenet.es [1] > > Antes de imprimir este correo electr?nico piense si es necesario hacerlo. > > El 16-12-2018 22:24, John Capo escribi?: > > On Thu, December 13, 2018 13:25, Egoitz Aurrekoetxea wrote: > Hi again! > > Else as a simplication way can you replicate any manner a 2.3 with some > newer version?. At least in manual mode (not rolling)?. > > The replication protocol in 2.4 is not compatible wit 2.3. > > There is no easy way. I'm rsync'ing one account at a time between 2.3 and 2.4 with IMAP/POP and LMTP access disabled, followed by a reconstruct, and then enabling IMAP/POP and LMTP access. > > Many terrabytes to go. > > You could also look at imapsync but its slow. > > Cheers. > > --- > > EGOITZ AURREKOETXEA > Departamento de sistemas > 944 209 470 > Parque Tecnol??gico. Edificio 103 > 48170 Zamudio (Bizkaia) > egoitz at sarenet.es www.sarenet.es [1] [1 [1]] Antes de imprimir este correo electr??nico piense si es > necesario hacerlo. > > El 13-12-2018 16:52, Egoitz Aurrekoetxea escribi??: > > Good afternoon, > > I was trying to upgrade part of our Cyrus imap installation, concretely that one consisting in > still 2.3. I was planning to set up Cyrus 3.0. I have seen all works properly except for the > unexpunge command because as someone stated here, a reconstruct -V max was needed.The problem > is that this reconstruct command, takes ages and I'm not able to keep the service offline so > many time. So I have been thinking in the following scenario : > > - Cyrus 2.3 master -> Cyrus 2.4 slave > > Get this 2.4 slave ready and set it as master. But here comes my first doubt. Does the 2.4 > replication work with the 2.3 replication?. Can in this pair, both (the 2.3 and the 2.4) be > both master and slave?. I mean to switch roles in the pair. Make one become master and the > other slave and vice versa?. > > Let's think now Cyrus 2.4 is ready and working. > > - Now, I would set up a new 3.0 slave. I know 2.4 could replicate with 3.0. So I would get the > 3.0 ready and then set 3.0 as master. Can in this pair both the 2.4 and 3.0 be master and > slave?. Meaning again to the same role switching commented before... to make one to be master > and the other slave or vice versa.... > > I'l will end up with 2 3.0 master and slave... but I need to trace the path... > > Does anyone see any other way?. > > Best regards, > > - > > -- > > EGOITZ AURREKOETXEA > Departamento de sistemas > 944 209 470 > Parque Tecnol??gico. Edificio 103 > 48170 Zamudio (Bizkaia) > egoitz at sarenet.es www.sarenet.es [1] [1 [1]] Antes de imprimir este correo electr??nico piense si es > necesario hacerlo. ---- > Cyrus Home Page: http://www.cyrusimap.org/ > List Archives/Info: http://lists.andrew.cmu.edu/pipermail/info-cyrus/ > To Unsubscribe: > https://lists.andrew.cmu.edu/mailman/listinfo/info-cyrus > > Links: > ------ > [1] http://www.sarenet.es Links: ------ [1] http://www.sarenet.es -------------- next part -------------- An HTML attachment was scrubbed... URL: